BACKGROUND: Thresholds of optimal thyroid status in old age are controversial. We investigated the longitudinal association between thyroid parameters and 10-year all-cause mortality risk in older outpatients with normal thyrotropin (TSH) and modification by sex and age.

METHODS: Baseline TSH, free thyroxine (fT4), and free triiodothyronine (fT3) were assessed in the Milan Geriatrics 75+ Cohort Study. 324 men and 609 women older than 75 years had normal TSH. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated for the associations between thyroid parameters and mortality risk using Cox regression. Sex-stratified analyses were adjusted for sociodemographic factors and comorbidities.

RESULTS: 233 men and 367 women died during follow-up. After adjustment, each 1-mU/L higher TSH was associated with decreased mortality risk in men (HR 0.83, 95% CI 0.69-0.98), but not in women (HR 1.09, 95% CI 0.95-1.24) (p for sex interaction = .006). Each 1-ng/L higher fT4 was associated with increased mortality risk in men (HR 1.11, 95% CI 1.02-1.22), but not in women (HR 0.98, 95% CI 0.93-1.04) (p for sex interaction = .013). Each 1-pg/mL higher fT3 was associated with decreased mortality risk in women (HR 0.77, 95% CI 0.60-0.98), but not in men (HR 0.80, 95% CI 0.57-1.13). The inverse association between TSH and mortality was most pronounced in men older than 85 years.

CONCLUSIONS: Among older outpatients with normal TSH, higher TSH and lower fT4 were associated with decreased mortality risk in men but not in women. When assessing thyroid status, sex and age should be taken into account.